Df2 nonlocal van der Waals

Developing Abinit
1

I an wondering the status for the full implementation of DF2 nonlocal van der Waals. This has been outstanding for years. Lots of materials need this. Thanks! Ron

2

Dear Ron,

I apologize for the prolonged implementation time. Currently vdW-DF1 and vdW-DF2 corrrections to the energy (non-local correlation) are working, however the corresponding potentials have bugs which prevent corrrect selfconsistency of the calculations. We are doing our best to pin-point the origin of the bug(s) and release this implementation as soon as possible. If you agree you can join me and perform some tests with the systems you are interested, in order to check this selfconsistency issue.

All the best,
Camilo

3

Hello both,

I was wondering which systems Ron was thinking of - for many layered materials Grimme corrections are enough (+ we have DFPT etc with those), and in some comparisons I saw there materials for which no simple functional works, and you have to go full RPA. What is the case for going specifically to DF-type functionals?

All the best

Matthieu